Engagement with and at School - OECD.org

2

Engagement with and at School This chapter examines several indicators of student engagement: arriving late for school, skipping days of school or classes, feeling a sense of belonging at school, and holding positive attitudes towards school. The chapter explores how these dispositions are associated with performance in mathematics, whether and how they are related to gender and socio-economic status, and how they have evolved among students since 2003.

READY TO LEARN: STUDENTS’ ENGAGEMENT, DRIVE AND SELF-BELIEFS – VOLUME III © OECD 2013

31

2

Engagement with and at School

It takes engagement and motivation to learn (Christenson, Reschly and Wylie, 2012; Wigfield et al., 2006). Students lose out on learning opportunities by skipping classes, arriving late or by being inattentive during lessons. Most of the students who participated in PISA 2012 hold positive views about education (see Figure III.2.1 for a comprehensive summary of the measures examined in this Volume). For example, 93% of students believe that trying hard at school is important and only 12% believe that school has been a waste of time. However many students are not engaged with school; they report being dissatisfied with school, not feeling in control of their ability to acquire knowledge, and not feeling capable of performing at high levels (see Skinner and Pitzer [2012] for a discussion of disengagement). Even more worryingly, specific subgroups of the student population are consistently at a high risk of suffering from low levels of engagement and motivation and of holding negative beliefs about their own capacities. This chapter, together with Chapters 3 and 4, examines variations in students’ drive to learn, their behaviours and dispositions towards school and their self-beliefs with regard to learning mathematics. These chapters identify the students who lack drive and motivation to succeed, who do not engage with school and learning, and who do not have confidence in their own abilities as learners. These students are at particular risk of not fulfilling their potential later on, either in the labour market or in their personal lives, because they are not engaged with learning when they are young.

What the data tell us • More than one in three students in OECD countries reported that they arrived late for school in the two weeks prior to the PISA test; and more than one in four students reported that they had skipped at least a class or a day of school during the same period. • On average across OECD countries, arriving late for school is associated with a 27-point lower score in mathematics, while skipping classes or days of school is associated with a 37-point lower score in mathematics – the equivalent of almost one full year of formal schooling. • Four out of five students in OECD countries agree or strongly agree that they feel happy at school. • Some 78% of disadvantaged students and 85% of advantaged students agree or strongly agree with the statement “I feel like I belong at school”.

Adolescence is a time when social acceptance, particularly by peers, can have a powerful influence on behaviour (Baumeister and Leary, 1995; Rubin, Bukowski and Parker., 2006). Peers can encourage and support students in their drive to achieve; they can also undermine students’ motivation and determination (Ladd et al., 2012). Because teenagers are particularly sensitive to peer pressures, students who are disengaged from school may be particularly at risk of developing conduct problems and associated negative outcomes (Barber, Stone and Eccles, 2010; Fredricks and Eccles, 2006; Griffiths et al., 2012; Juvonen, Espinoza and Knifsend, 2012). For many students, school is essential to their longterm well-being; this is reflected in their participation in academic and non-academic activities organised by their school. While a large majority of students tends to have good relations with school staff and other students and feel that they belong at school, others do not share this sense of belonging. This latter group may, in the long run, become disaffected with school and, as a result, have poorer outcomes (Finn, 1989; Jenkins, 1995; Due et al., 2003; Bonell, Fletcher and McCambridge, 2007). The social aspects of engagement at school are manifested in students’ willingness to work with others, and their ability to function in and contribute to social institutions. When students feel a sense of belonging at school, their engagement is often enhanced (Juvonen, Espinoza and Knifsend, 2012); when they don’t, behavioural problems often follow. When families and education systems fail to address these problems when children are still in school, these problems – and their repercussions – are likely to follow students into adulthood (Offord and Bennett, 1994; Bennett and Offord, 2001). Disruptive behaviour, poor attendance at and negative dispositions towards school are associated with low academic performance and are related to such negative outcomes as low levels of emotional well-being, school dropout, delinquency and drug abuse (e.g. Valeski and Stipek, 2001; Baker, Sigmon and Nugent, 2001; Lee and Burkam, 2003; McCluskey, Bynum and Patchin, 2004).

32

© OECD 2013 READY TO LEARN: STUDENTS’ ENGAGEMENT, DRIVE AND SELF-BELIEFS – VOLUME III


2

• Figure III.2.1 • How PISA 2012 measures students’ engagement with and at school

Lack of punctuality Students’ reports on whether they arrived late for school in the two weeks before the test

Absenteeism Students’ reports on whether they skipped classes or days of school in the two weeks before the test

Sense of belonging Derived index based on students’ reports about their feelings of social connectedness, happiness and satisfaction at school

Attitudes towards school (learning outcomes and learning activities) Derived indices based on students’ reports about the importance of school for their future and about the importance of and pleasure they derive from working hard at school

LACK OF PUNCTUALITY: ARRIVING LATE FOR SCHOOL PISA reveals that, in 2012, significant proportions of students arrived late for school, without authorisation, at least once in the two weeks prior to the PISA test. On average across OECD countries, more than one in three students (35%) reported having arrived late at least once during that two-week period. Figure III.2.2 shows that 25% of students arrived late once or twice, 6% arrived late three or four times, and 4% arrived late five times or more. Figure III.2.2 also reveals how lack of punctuality is particularly acute in Uruguay, Bulgaria, Costa Rica, Latvia, Sweden, Portugal, Israel, Chile, Peru and Tunisia. In these countries, more than half of students reported arriving late at least once in the two weeks prior to the PISA test. By contrast, in Japan and Hong Kong-China less than 15% of students reported having arrived late for school (Table III.2.1a). Between-country variations in the proportion of students who arrive late for school mask large within-country differences across different subgroups of the population. In 28 countries and economies socio-economically disadvantaged students were more likely to report having arrived late for school than socio-economically advantaged students. Lack of punctuality is a widespread phenomenon that is particularly acute among socio-economically disadvantaged students in many countries and economies. In Bulgaria, Sweden, Israel, Chile and New Zealand more than one in two socio-economically disadvantaged students reported having arrived late for school at least once in the two weeks before the PISA test, and they were significantly more likely to report having arrived late than their advantaged peers. Lack of punctuality is strongly associated with students’ socio-economic status for a variety of reasons. In some countries and geographical locations, disadvantaged students – who are students in the bottom quarter of the PISA index of economic, social and cultural status – may be more likely to rely on public transportation. In others, they may live in neighbourhoods that are not well-served by efficient public transportation systems and their families may not have access to private means of transportation. They may also live in neighbourhoods that are comparatively less safe. The parents of disadvantaged students may also struggle to meet multiple demands on their time and so may not be able to keep track of their child’s punctuality. These students may also have to help around the house or work for pay to help support their families.


33

2


• Figure III.2.2 • Percentage of students who arrive late for school Percentage of students who reported to arrive late for school in the two weeks prior to the PISA test None

One or two times

Three or four times

Five times or more

Japan Hong Kong-China Viet Nam Shanghai-China Liechtenstein Singapore Austria Chinese Taipei Germany Hungary Switzerland Korea Macao-China Slovak Republic Indonesia Czech Republic Belgium Ireland Kazakhstan Luxembourg Norway United States Netherlands United Arab Emirates United Kingdom France Malaysia Brazil Croatia Thailand Iceland Italy OECD average Albania Spain Jordan Australia Colombia Denmark Montenegro Qatar Slovenia Mexico Estonia Serbia New Zealand Poland Finland Canada Lithuania Turkey Romania Russian Federation Argentina Greece Tunisia Peru Chile Israel Portugal Sweden Latvia Costa Rica Bulgaria Uruguay

100 90

Japan Hong Kong-China Viet Nam Shanghai-China Liechtenstein Singapore Austria Chinese Taipei Germany Hungary Switzerland Korea Macao-China Slovak Republic Indonesia Czech Republic Belgium Ireland Kazakhstan Luxembourg Norway United States Netherlands United Arab Emirates United Kingdom France Malaysia Brazil Croatia Thailand Iceland Italy OECD average Albania Spain Jordan Australia Colombia Denmark Montenegro Qatar Slovenia Mexico Estonia Serbia New Zealand Poland Finland Canada Lithuania Turkey Romania Russian Federation Argentina Greece Tunisia Peru Chile Israel Portugal Sweden Latvia Costa Rica Bulgaria Uruguay

80

70

60

50

40

30

20

10

0

10

20

30

40

50

60

70

80

90 100

Percentage of students

Countries and economies are ranked in descending order of the percentage of students who had never arrived late for school in the two weeks prior to the PISA test. Source: OECD, PISA 2012 Database, Table III.2.1a. 1 2 http://dx.doi.org/10.1787/888932963806

34



2

In 32 countries and economies, girls were less likely than boys to report having arrived late for school in the two weeks before the PISA test. Although the difference in the proportion of boys and girls who reported having arrived late is small – 2%, on average across OECD countries – it is larger than ten percentage points in Lithuania, Thailand and Poland. • Figure III.2.3 • Socio‑economic disparities in arriving late for school Students in bottom quarter of ESCS Students in top quarter of ESCS

60 50 40 30 20 10 0

Bulgaria Sweden Uruguay Israel Portugal Chile Latvia Costa Rica New Zealand Peru Tunisia Romania Russian Federation Argentina Finland Canada Turkey Lithuania Greece Estonia Denmark Slovenia Qatar Serbia United States Spain Montenegro Australia Poland Italy OECD average France Iceland Malaysia United Kingdom Jordan Colombia Netherlands Mexico Thailand United Arab Emirates Hungary Kazakhstan Luxembourg Norway Croatia Czech Republic Brazil Ireland Belgium Slovak Republic Korea Macao-China Singapore Chinese Taipei Germany Switzerland Indonesia Liechtenstein Austria Viet Nam Shanghai-China Hong Kong-China Japan

(439) (478) (409) (466) (487) (423) (491) (407) (500) (368) (388) (445) (482) (388) (519) (518) (448) (479) (453) (521) (500) (501) (376) (449) (481) (484) (410) (504) (518) (485) (494) (495) (493) (421) (494) (386) (376) (523) (413) (427) (434) (477) (432) (490) (489) (471) (499) (391) (501) (515) (482) (554) (538) (573) (560) (514) (531) (375) (535) (506) (511) (613) (561) (536)

Percentage of students who reported having arrived late for school

70

Notes: ESCS refers to the PISA index of economic, social and cultural status. Differences that are statistically significant at the 5% level (p < 0.05) are marked in a darker tone. Mean mathematics performance is shown above the country name in parenthesis. Countries and economies are ranked in descending order of the percentage of socio-economically disadvantaged students who reported having arrived late for school in the two weeks prior to the PISA test. Source: OECD, PISA 2012 Database, Tables I.2.3a and III.2.1a. 1 2 http://dx.doi.org/10.1787/888932963806

Overall students’ punctuality has improved over the past nine years. In 2003, an average of 36% of students in OECD countries with comparable data between 2003 and 2012 reported having arrived late at least once during the two weeks prior to the PISA test; in 2012, and among these same countries, this percentage decreased to 34%. Fifteen countries and economies saw significant improvements in punctuality, and the share of students arriving late shrank by more than five percentage points in Mexico, Spain, Norway, Luxembourg, Japan, Indonesia, Italy, Iceland and the Netherlands. By contrast, nine countries and economies saw an increase in the percentage of students arriving late. This increase was greater than five percentage points in Poland, Macao-China, Latvia and the Russian Federation and greater than ten percentage points in Tunisia and Turkey (Figure III.2.4). On average across OECD countries, boys and girls, as well as advantaged and disadvantaged students, were less likely in 2012 than in 2003 to report having arrived late. Still, the improvement was greater among girls than among boys, and among socio-economically advantaged students than among disadvantaged students. Trends between 2003 and 2012 show better punctuality among girls than boys in Turkey, Denmark and Korea, where the gender gap in punctuality widened by around five percentage points or more, in favour of girls. In Korea in 2003, girls were more likely than boys to arrive late; by 2012, girls and boys were similarly punctual. In Turkey, boys and girls in 2003 reported having arrived late at a similar rate; but by 2012, boys were eight percentage points more likely than girls to arrive late. Trends also show increasingly better punctuality among advantaged students than disadvantaged students in five countries and economies. In Luxembourg, for example, advantaged students were eleven percentage points less likely to report having arrived late in 2012 than they were in 2003, the trend among students in the bottom quarter of that index signals no improvement in punctuality during the period (Figure III.2.5). READY TO LEARN: STUDENTS’ ENGAGEMENT, DRIVE AND SELF-BELIEFS – VOLUME III © OECD 2013

35

2


• Figure III.2.4 • Change between 2003 and 2012 in students arriving late for school

15 10 5 0 -5 -10 -15

Turkey

Latvia

Tunisia

Macao-China

Poland

Russian Federation

Sweden

Czech Republic

Uruguay

Slovak Republic

Germany

Greece

Portugal

France

Thailand

Canada

Belgium

Ireland

Australia

Korea

Finland

OECD average 2003

Austria

Liechtenstein

Switzerland

Brazil

Hong Kong-China

Hungary

New Zealand

Denmark

United States

Spain

Mexico

Norway

Japan

Luxembourg

Italy

Iceland

Indonesia

-20 Netherlands

Change in the percentage of students who arrived late at least once in the two weeks prior to the PISA assessment

20

Notes: Statistically significant changes at the 5% level (p < 0.05) between PISA 2003 and 2012 are marked in a darker tone. Only countries and economies with comparable data from PISA 2003 and PISA 2012 are shown. The OECD average compares only OECD countries with comparable results for students arriving late since 2003. Countries and economies are ranked in ascending order of the change in the percentage of students who reported having arrived late at least once in the two weeks prior to the PISA test between PISA 2003 and PISA 2012. Source: OECD, PISA 2012 Database, Table III.2.1b. 1 2 http://dx.doi.org/10.1787/888932963806

• Figure III.2.5 • Change between 2003 and 2012 in socio‑economic disparities in arriving late for school

Advantaged students have become more likely to arrive late than disadvantaged students

5

0

-5

-10

Portugal

New Zealand

Iceland

Luxembourg

Austria

Tunisia

Spain

United States

Finland

Sweden

Slovak Republic

France

Czech Republic

Canada

Denmark

Korea

Hungary

Australia

Russian Federation

Norway

Switzerland

Macao-China

Uruguay

Germany

Turkey

Hong Kong-China

Brazil

Ireland

Japan

Italy

Poland

Netherlands

Greece

Thailand

Latvia

Mexico

Belgium

-15

OECD average 2003

Disadvantaged students have become more likely to arrive late than advantaged students Indonesia

Change in the percentage-point difference in socio-economic disparities (advantaged-disadvantaged) in the percentage of students arriving late for school

10

Notes: Statistically significant percentage point changes at the 5% level (p < 0.05) are marked in a darker tone. Advantaged/disadvantaged students are students in the top/bottom quarter of the PISA index of economic, social and cultural status. Only countries and economies with comparable data from PISA 2003 and PISA 2012 are shown. The OECD average compares only OECD countries with comparable results for students arriving late since 2003. Countries and economies are ranked in descending order of the change in socio-economic disparities in the percentage of students who reported having arrived late at least once in the two weeks prior to the PISA test between 2003 and 2012. Source: OECD, PISA 2012 Database, Table III.2.1b. 1 2 http://dx.doi.org/10.1787/888932963806

36



2

Figure III.2.6 shows that students who reported having arrived late for school at least once in the two weeks before the PISA test have lower scores than students who reported that they had not arrived late for school during that period (Table III.2.1a). Across OECD countries, the difference in mathematics performance that is associated with arriving late for school is 27 score points. On average, students who reported that they did not arrive late for school score 504 points while those who reported arriving late for school score 477 points. All countries and economies except for Greece, Albania, Costa Rica and Tunisia show a performance gap associated with arriving late for school. In Hungary, Chinese Taipei, Singapore and Macao-China, the difference in performance between students who reported having arrived late for school and those who did not is 45 score points or more (Table III.2.1c). The last section of this chapter and Chapter 7 illustrate in detail the extent to which differences in socio-economic status explain part of the relationship between students’ lack of punctuality and mathematics performance. Chapter 5 of Volume IV of this series examines students’ lack of punctuality and truancy as two of the factors that determine school climate. • Figure III.2.6 • Relationship between arriving late for school and mathematics performance Average student 10th percentile (lowest-achieving students) 90th percentile (highest-achieving students) Score point difference in mathematics, associated with arriving late for school

20 10 0 -10 -20 -30 -40 -50 -60

Hungary Chinese Taipei Singapore Macao-China Korea Shanghai-China Qatar New Zealand Hong Kong-China Malaysia United Kingdom United States France Belgium Norway Netherlands Japan Liechtenstein Canada Czech Republic Viet Nam Australia Italy Iceland Slovak Republic Sweden Ireland United Arab Emirates Bulgaria Spain Finland OECD average Russian Federation Argentina Chile Slovenia Croatia Thailand Peru Denmark Luxembourg Estonia Lithuania Poland Israel Germany Serbia Indonesia Turkey Kazakhstan Montenegro Portugal Mexico Austria Jordan Romania Latvia Colombia Uruguay Switzerland Brazil Tunisia Greece Costa Rica Albania

-70

Note: Differences that are statistically significant at the 5% level (p < 0.05) are marked in a darker tone. Countries and economies are ranked in ascending order of the average score point difference in mathematics that is associated with students arriving late for school. Source: OECD, PISA 2012 Database, Table III.2.1c. 1 2 http://dx.doi.org/10.1787/888932963806

The findings presented in Figure III.2.6 suggest that performance differences associated with a lack of punctuality are particularly strong at the bottom of the performance distribution. On average across OECD countries, the gap in scores associated with arriving late for school is 31 points among the lowest-achieving students and 20 points among the highest-achieving students (Table III.2.1c).1 This average masks large differences across countries, however. In 24 countries and economies, the difference in performance associated with arriving late for school is larger than 10 score points at the bottom than at the top tail of the performance distribution. In Japan, Hong Kong-China and Austria the performance gap between the most and least able students is at least 30 score points. In 12 countries and economies, the lowest-achieving students who reported having arrived late for school show poorer performance in mathematics than the lowest-achieving students who did not report having arrived late for school, but no such difference can be observed among the highest-achieving students.


37

2


Box III.2.1. The cyclical nature of the relationship between students’ dispositions, behaviours and self‑beliefs and mathematics performance Students who hold positive dispositions towards school, who are motivated to learn mathematics and who have a positive image of themselves as mathematics learners perform better in the PISA mathematics assessment. However, this finding cannot be interpreted as direct evidence of a causal relationship between students’ dispositions, behaviours and self-beliefs and achieving high levels of mathematics proficiency. Rather, evidence presented in this chapter reflects the cumulative observed association between students’ dispositions, behaviours and self-beliefs and how good they are in mathematics. What does cumulative association mean? Studies in education and applied psychology suggest that mathematics proficiency is the result of multiple developmental cycles. Students’ dispositions towards mathematics and learning, motivation, engagement in mathematics activities and mathematics proficiency are mutually reinforcing. Positive reinforcement operates at two levels. The first reflects the fact that the future depends on the past. Past behaviours matter for current and future behaviours and past mathematics performance is also a very good predictor of future mathematics performance (Fredericks, Blumenfeld and Paris, 2004; Baumert, Nagy and Lehmann, 2012). This suggests that a student’s past dispositions, behaviours and self-beliefs will influence his or her future dispositions, behaviours and self-beliefs. The second level indicates that associations among dispositions, behaviours and self-beliefs and performance are circular. Students’ dispositions, behaviours and self-beliefs and mathematics performance are mutually dependent. For example, students who believe they can solve mathematics problems become better at solving them; and when they see that they are good at mathematics and expect good performance in mathematics, they tend to have higher levels of self-efficacy, to enjoy mathematics and to engage with school and mathematics (Nurmi et al., 2003). The graph below illustrates how results on associations between students’ dispositions, behaviours and self-beliefs and how well they do in mathematics should be interpreted in the context of the two levels of reinforcement. The cumulative relationship between mathematics performance and student engagement, drive, motivation and self-beliefs Engagement/ Drive/ Self-beliefs

Time

Engagement/ Drive/ Self-beliefs Engagement/ Drive/ Self-beliefs Engagement/ Drive/ Self-beliefs

Performance

Performance

Performance

Performance

The evidence that emerges from PISA on the positive interplay between students’ dispositions, behaviours, selfbeliefs and performance in mathematics suggests that promoting proficiency in mathematics and promoting a passion for mathematics, school and learning does not necessarily involve trade-offs. Students who are highly engaged and are effective learners are most likely to be proficient in mathematics and students who are proficient in mathematics are also those students who hold positive dispositions towards schools and learning, who attend school regularly and who have positive self-beliefs about mathematics.

...

38



2

Sources: Baumert, J., J. Nagy and R. Lehmann (2012), “Cumulative Advantages and the Emergence of Social and Ethnic Inequality: Matthew Effects in Reading and Mathematics Development Within Elementary Schools?”, Child Development, Vol. 83, No. 4, pp. 1347-1367. Fredricks, J.A., P.C. Blumenfeld and A.H. Paris (2004), “School engagement: Potential of the Concept, State of the evidence”, Review of Educational Research, Vol. 74, pp. 59-109. Nurmi, J.E. et al. (2003), “The Role of Success expectation and task-avoidance in Academic Performance and Satisfaction: three Studies on Antecedents, Consequences and Correlates”, Contemporary Education Psychology, Vol. 28, pp. 59-90.

ABSENTEEISM: SKIPPING CLASSES OR DAYS OF SCHOOL Regular absenteeism represents a missed learning opportunity, signifies lack of interest, and also has negative consequences on students’ classmates because it contributes to a disruptive learning environment. Students who took part in PISA 2012 were asked to report how many times they skipped classes or days of school without authorisation in the two weeks prior to the PISA assessment. Results presented in Figures III.2.7 and III.2.8 reveal that absenteeism is a problem in many countries. Across OECD countries, 18% of students reported that they had skipped at least one class and 15% reported that they had skipped at least an entire day of school without authorisation in the two weeks before the PISA test. In Argentina, Turkey, Italy and Jordan, 40% of students or more reported that they had skipped at least one day of school, and in Latvia, Turkey, Argentina, Romania, Costa Rica and Greece, 40% of students or more reported that they had skipped at least one class. In Latvia, Turkey, Argentina, Greece and Romania, 4% of students or more reported having skipped a class five times or more in the previous two weeks, and in Turkey and Argentina, more than 7% of students reported having skipped five or more days of school during that period (Tables III.2.2a and III.2.2b) Figures III.2.9 and III.2.10 illustrate differences in the proportion of socio-economically advantaged and disadvantaged students who reported having skipped classes or days of school. In many countries skipping classes or days of school is a particularly acute problem among disadvantaged students: across OECD countries, the difference between advantaged and disadvantaged students who reported having skipped classes is two percentage points while in having skipped days of school it is six percentage points. Across OECD countries, 19% of disadvantaged students (compared with 17% of advantaged students) reported having skipped classes, while 18% of disadvantaged students (compared with 12% of advantaged students) reported having skipped days of school (Tables III.2.2a and III.2.2b). Figure III.2.11 shows that students who reported having skipped classes or days of school at least once in the two weeks prior to the PISA test have lower scores than students who reported not skipping classes or days of school (Table III.2.2c). Across OECD countries, the difference in mathematics performance that is associated with skipping classes or days of school is 37 score points; in Korea, Japan and Chinese Taipei that difference is 80 score points or more. In every country except Brazil, Colombia and Israel, skipping classes or days of school is associated with a performance disadvantage. On average, the difference in performance associated with skipping classes or days of school tends to be similar between the most and the least able students. The OECD average masks large differences across countries, however. For example, in 20 countries and economies, this difference is at least 10 score points larger among the least able students than among the most able students. In Peru, New Zealand, Serbia, Croatia and Bulgaria, skipping classes is more negatively associated with performance among the most able students than among the least able students and this difference is at least 10 score points. Students’ absenteeism and lack of punctuality can have a disruptive effect on classes and schools more broadly. Chapter 5 of this volume identifies the extent to which students who reported having arrived late or having skipped classes or days of school are concentrated in particular schools. That the lowest-achieving students tend to suffer the most from arriving late for school and, in many countries and economies, from skipping classes or days of school may reflect the fact that these are the students who need – and should take advantage of – learning opportunities the most.


39

2


• Figure III.2.7 • Percentage of students who skip classes Percentage of students who reported that they skipped classes in the two weeks prior to the PISA test None

One or two times

Three or four times

Five times or more

Japan Korea Hong Kong-China Shanghai-China Liechtenstein Macao-China Viet Nam Luxembourg Czech Republic Belgium Hungary Chinese Taipei Germany Switzerland Netherlands Iceland Slovak Republic Norway United Kingdom Peru Ireland Singapore Austria United States Australia New Zealand Chile Finland Colombia Denmark France Kazakhstan OECD average Brazil Albania Qatar Poland Sweden Mexico United Arab Emirates Croatia Uruguay Canada Indonesia Malaysia Tunisia Slovenia Thailand Serbia Portugal Jordan Estonia Russian Federation Israel Montenegro Spain Lithuania Bulgaria Italy Greece Costa Rica Romania Argentina Turkey Latvia

100 90

Japan Korea Hong Kong-China Shanghai-China Liechtenstein Macao-China Viet Nam Luxembourg Czech Republic Belgium Hungary Chinese Taipei Germany Switzerland Netherlands Iceland Slovak Republic Norway United Kingdom Peru Ireland Singapore Austria United States Australia New Zealand Chile Finland Colombia Denmark France Kazakhstan OECD average Brazil Albania Qatar Poland Sweden Mexico United Arab Emirates Croatia Uruguay Canada Indonesia Malaysia Tunisia Slovenia Thailand Serbia Portugal Jordan Estonia Russian Federation Israel Montenegro Spain Lithuania Bulgaria Italy Greece Costa Rica Romania Argentina Turkey Latvia

80

70

60

50

40

30

20

10

0

10

20

30

40

50

60

70

80

90 100


Countries and economies are ranked in descending order of the percentage of students who reported not having skipped classes in the two weeks prior to the PISA test. Source: OECD, PISA 2012 Database, Table III.2.2a. 1 2 http://dx.doi.org/10.1787/888932963806

40



2

• Figure III.2.8 • Percentage of students who skip days of school Percentage of students who reported that they skipped days of school in the two weeks prior to the PISA test None

One or two times

Three or four times

Five times or more

Shanghai-China Japan Korea Liechtenstein Iceland Netherlands Hong Kong-China Ireland Chinese Taipei Colombia Macao-China Switzerland Germany Belgium Czech Republic Hungary Luxembourg Norway Sweden Chile Austria Viet Nam Slovak Republic France Denmark Finland Indonesia Croatia Serbia Slovenia Peru Singapore OECD average Albania Estonia Poland Qatar New Zealand United Kingdom Thailand Lithuania Portugal Kazakhstan Brazil Tunisia Mexico United States Russian Federation Greece Canada Latvia Uruguay Montenegro Bulgaria Spain Malaysia Israel Costa Rica Australia Romania United Arab Emirates Jordan Italy Turkey Argentina

100 90

Shanghai-China Japan Korea Liechtenstein Iceland Netherlands Hong Kong-China Ireland Chinese Taipei Colombia Macao-China Switzerland Germany Belgium Czech Republic Hungary Luxembourg Norway Sweden Chile Austria Viet Nam Slovak Republic France Denmark Finland Indonesia Croatia Serbia Slovenia Peru Singapore OECD average Albania Estonia Poland Qatar New Zealand United Kingdom Thailand Lithuania Portugal Kazakhstan Brazil Tunisia Mexico United States Russian Federation Greece Canada Latvia Uruguay Montenegro Bulgaria Spain Malaysia Israel Costa Rica Australia Romania United Arab Emirates Jordan Italy Turkey Argentina

80

70

60

50

40

30

20

10

0

10

20

30

40

50

60

70

80

90 100


Countries and economies are ranked in descending order of the percentage of students who reported not having skipped days of school in the two weeks prior to the PISA test. Source: OECD, PISA 2012 Database, Table III.2.2b. 1 2 http://dx.doi.org/10.1787/888932963806


41

Argentina (388) Italy (485) Turkey (448) Jordan (386) United Arab Emirates (434) Romania (445) Australia (504) Spain (484) Bulgaria (439) Israel (466) Costa Rica (407) Malaysia (421) Latvia (491) Uruguay (409) New Zealand (500) Russian Federation (482) United States (481) Kazakhstan (432) Lithuania (479) Montenegro (410) Canada (518) Portugal (487) Greece (453) Tunisia (388) Brazil (391) United Kingdom (494) Poland (518) Mexico (413) Thailand (427) OECD average (494) Singapore (573) Peru (368) Slovenia (501) Estonia (521) Croatia (471) Viet Nam (511) France (495) Slovak Republic (482) Denmark (500) Finland (519) Qatar (376) Serbia (449) Indonesia (375) Sweden (478) Hungary (477) Luxembourg (490) Austria (506) Chile (423) Norway (489) Czech Republic (499) Chinese Taipei (560) Belgium (515) Germany (514) Switzerland (531) Macao-China (538) Ireland (501) Liechtenstein (535) Colombia (376) Hong Kong-China (561) Netherlands (523) Korea (554) Iceland (493) Japan (536) Shanghai-China (613)

Percentage of students skipping days of school

Latvia (491) Romania (445) Argentina (388) Turkey (448) Costa Rica (407) Bulgaria (439) Greece (453) Lithuania (479) Spain (484) Italy (485) Russian Federation (482) Estonia (521) Portugal (487) Montenegro (410) Slovenia (501) Canada (518) Israel (466) Serbia (449) Uruguay (409) Sweden (478) Thailand (427) Jordan (386) Tunisia (388) Croatia (471) Indonesia (375) United Arab Emirates (434) Malaysia (421) Kazakhstan (432) New Zealand (500) Poland (518) Qatar (376) OECD average (494) Chile (423) France (495) Finland (519) Denmark (500) Brazil (391) Slovak Republic (482) United States (481) Mexico (413) Australia (504) Iceland (493) Peru (368) Norway (489) Hungary (477) Colombia (376) Chinese Taipei (560) United Kingdom (494) Singapore (573) Ireland (501) Austria (506) Belgium (515) Germany (514) Netherlands (523) Luxembourg (490) Switzerland (531) Liechtenstein (535) Czech Republic (499) Viet Nam (511) Macao-China (538) Korea (554) Japan (536) Shanghai-China (613) Hong Kong-China (561)

Percentage of students skipping classes

2

42 Engagement with and at School

• Figure III.2.9 • Socio‑economic disparities in skipping classes

70

70

Students in bottom quarter of ESCS Students in top quarter of ESCS

60

50

40

30

20

10

0

Notes: ESCS refers to the PISA index of economic, social and cultural status. Differences that are statistically significant at the 5% level (p < 0.05) are marked in a darker tone. Countries and economies are ranked in descending order of the difference between the percentage of students skipping classes who are in the bottom quarter of ESCS and those who are in the top quarter of ESCS (top – bottom). Source: OECD, PISA 2012 Database, Tables I.2.3a and III.2.2a. 1 2 http://dx.doi.org/10.1787/888932963806

• Figure III.2.10 • Socio‑economic disparities in skipping days of school

Students in bottom quarter of ESCS Students in top quarter of ESCS

60

50

40

30

20

10

0

Notes: ESCS refers to the PISA index of economic, social and cultural status. Differences that are statistically significant at the 5% level (p < 0.05) are marked in a darker tone. Countries and economies are ranked in descending order of the difference between the percentage of students skipping days of school who are in the bottom quarter of ESCS and those who are in the top quarter of ESCS (top – bottom). Source: OECD, PISA 2012 Database, Tables I.2.3a and III.2.2b. 1 2 http://dx.doi.org/10.1787/888932963806



2

• Figure III.2.11 • The relationship between skipping classes and days of school and mathematics performance Average student 10th percentile (lowest-achieving performers) 90th percentile (highest-achieving performers)

0 -20 -40 -60 -80 -100 -120 Korea Chinese Taipei Japan New Zealand Belgium Hong Kong-China Hungary Liechtenstein Norway Luxembourg Viet Nam Macao-China Iceland Croatia Sweden Bulgaria Slovak Republic Slovenia Lithuania Peru Australia Estonia OECD average Finland Denmark Spain United Kingdom Czech Republic Shanghai-China France Portugal Italy Poland Chile Canada United Arab Emirates Russian Federation Singapore Argentina Kazakhstan Switzerland United States Malaysia Germany Serbia Uruguay Thailand Romania Indonesia Qatar Austria Greece Ireland Montenegro Tunisia Latvia Jordan Mexico Netherlands Costa Rica Colombia Israel Brazil Albania Turkey

Score point difference in mathematics, associated with skipping classes or days of school

20

Note: Differences that are statistically significant at the 5% level (p < 0.05) are marked in a darker tone. Countries and economies are ranked in ascending order of the average score-point difference in mathematics that is associated with students skipping classes or days of school. Source: OECD, PISA 2012 Database, Table III.2.2c. 1 2 http://dx.doi.org/10.1787/888932963806

SENSE OF BELONGING For young children, the family is the centre of their social and emotional world. That changes during adolescence. Teenagers begin to look farther afield for support and acceptance (Baumeister and Leary, 1995); and often that acceptance (or lack of it) has a strong impact on adolescents’ sense of self-worth (Harter, 1999). Rejection by one’s peers can be a hurtful – indeed sometimes physically painful – experience (Eisenberger, Lieberman and Williams, 2003; Eisenberger and Lieberman, 2006; Kross et al., 2011). Indicators of social connectedness can show the extent to which families, schools and education systems foster overall student well-being. A sense of belonging reflects how connected students feel with their school and peers. Students tend to thrive when they form positive relationships with peers, feel part of a social group, and feel at ease at school. A lack of connectedness can adversely affect students’ perceptions of themselves, their satisfaction with life, and their willingness to learn and to put effort into their studies. In 2012, as in 2003, PISA asked students to report whether they “strongly agree”, “agree”, “disagree” or “strongly disagree” that they feel like an outsider or left out of things, that they make friends easily, that they feel like they belong, that they feel awkward and out of place, that other students seem to like them, or that they feel lonely. For the first time, PISA 2012 asked students to evaluate their happiness at, and satisfaction with, school and to reflect on whether their school environment approaches their idea of an ideal situation. As schools are a, if not the, primary social environment for 15-year-olds, these subjective evaluations provide a good indication of whether education systems are able to foster or hinder overall student well-being. Student responses to these nine questions were used to construct the index of sense of belonging, which was standardised to have a mean of 0 and a standard deviation of 1 across OECD countries. As Figure III.2.12 and Table III.2.3a show, across OECD countries, 81% of students feel that they belong, 87% of students agree or strongly agree that they can make friends easily, and 89% of students disagree that they feel like an outsider or feel left out of things. Some 80% of students feel happy at school, 78% are satisfied with school, and 61% believe that conditions are ideal in their school (Table III.2.3a). However, in some countries sizable minorities of students do not have positive relationships with their peers at school, do not feel connected with their school, and are not happy or satisfied with school. Worryingly, in many countries, students’ sense of belonging deteriorated somewhat between 2003 and 2012.


43

2


• Figure III.2.12 • Students’ sense of belonging Percentage of students who reported “agree” or “strongly agree” (a) or who reported “disagree” or “strongly disagree” (b) % 100

80

60

40

20

0 I feel like an outsider (or left out of things) at school b

I make friends easily at school a

I feel like I belong at school a

I feel awkward Other students and out of place seem to in my school b like me a

I feel lonely at school b

I feel happy Things are ideal I am satisfied at school a in my school a with my school a

Source: OECD, PISA 2012 Database, Table III.2.3a. 1 2 http://dx.doi.org/10.1787/888932963806

• Figure III.2.13 • Change between 2003 and 2012 in students’ sense of belonging Change in the mean index of sense of belonging between 2003 and 2012

0.6 0.5 0.4 0.3 0.2 0.1 0 -0.1 -0.2 -0.3

Notes: Statistically significant changes at the 5% level (p < 0.05) between PISA 2003 and 2012 are marked in a darker tone. Only countries and economies with comparable data from PISA 2003 and PISA 2012 are shown. OECD average compares only OECD countries with comparable indices of sense of belonging since 2003. Countries and economies are ranked in descending order of the change in the mean index of sense of belonging between PISA 2003 and PISA 2012. Source: OECD, PISA 2012 Database, Table III.2.3f. 1 2 http://dx.doi.org/10.1787/888932963806

44


Brazil

Italy

Sweden

Finland

Australia

Poland

Greece

Norway

Slovak Republic

New Zealand

Ireland

Canada

Tunisia

Czech Republic

Portugal

Uruguay

Denmark

Luxembourg

Latvia

OECD average 2003

Mexico

Netherlands

Hungary

Korea

Germany

France

Macao-China

Austria

Russian Federation

Thailand

Hong Kong-China

Spain

Iceland

Belgium

Switzerland

Japan

Indonesia

Turkey

Liechtenstein

-0.4


2

Although students’ sense of belonging is generally strong across participating countries and economies, it weakened slightly, on average across OECD countries. In 2003, 93% of students did not feel like outsiders or left out of things at school, but by 2012, that proportion shrank somewhat to 89%. In 31 of the 38 countries and economies with comparable data between 2003 and 2012, the share of students who reported feeling like an outsider increased significantly. Most notably, in Tunisia, Thailand and France the share of students who agreed that they feel like an outsider at school grew by more than ten percentage points between 2003 and 2012. The share of students who reported feeling awkward and out place at school increased significantly in 26 countries and economies with comparable data between 2003 and 2012, and grew by more than five percentage points in ten countries and economies during the period. A weakening of students’ sense of belonging between 2003 and 2012 is particularly notable in Brazil, Sweden, Italy, Finland and Australia. In Sweden, for example, there were consistently more students in 2012 than in 2003 who reported feeling like an outsider, awkward and out of place in school. In Australia, the share of students who reported that they feel like they belong at school shrank by around ten percentage points and, in Brazil, the share of students who reported feeling lonely at school increased by more than ten percentage points (Figure III.2.13).

Box III.2.2. Interpreting PISA indices Indices used to characterise students’ dispositions, behaviours and self-beliefs were constructed so that the average OECD student would have an index value of zero and about two-thirds of the OECD student population would be between the values of -1 and 1 (i.e. the index has a standard deviation of 1). Negative values on the index, therefore, do not imply that students responded negatively to the underlying question. Rather, students with negative scores are students who responded less positively than the average response across OECD countries. Likewise, students with positive scores are students who responded more positively than the average student in the OECD area (see Annex A3 for a detailed description of how indices were constructed). Most of the indicators of engagement, drive and self-beliefs are based on students’ self-reports. Such measures can thus suffer from a degree of measurement error because students are asked to assess their engagement, drive and self-beliefs retrospectively. Apart from potential measurement error, cultural differences in attitudes towards self-enhancement can influence country-level results in students’ self-reported engagement, drive and self-beliefs (Bempechat, Jimenez and Boulay, 2002). The literature consistently shows that response biases, such as social desirability, acquiescence and extreme response choice, are more common in countries with low GDP than in more affluent countries, as they are, within countries, among individuals with lower socio-economic background and less education (Buckley, 2009). As in the 2003 PISA cycle, in 2012 many of the self-reported indicators of engagement, drive and self-beliefs are strongly and positively associated with mathematics performance within countries, but show a weak or negative association with performance between countries. This may be due to different response biases across countries or the fact that country-level differences in mathematics performance are due to many factors that go beyond levels of engagement, drive and self-beliefs, and that are negatively associated with mathematics performance and positively associated with engagement, drive and self-beliefs. In PISA 2012 new survey methods were introduced to enhance the validity of questionnaire indexes, especially for cross-country comparisons. One of the new methods introduced is an alternative scoring of Likert-type items based on so-called anchoring vignettes (King and Wand, 2007). Annex A6 contains a full description of the anchoring vignettes methodology. Caution is advised when comparing levels of engagement, drive and self-beliefs across countries because different students, particularly students in different countries, may not always mean the same thing when answering questions. The PISA 2012 Technical Report (OECD, forthcoming) contains a detailed description of all the steps that were taken in PISA 2012 to ensure the highest possible level of cross-country comparability and to assess the validity of cross-country comparisons based on the indices featured in the report.* PISA scale indices, like the PISA index of economic, social and cultural status, index of sense of belonging, index of attitudes towards school, index of intrinsic motivation to learn mathematics, index of instrumental motivation to learn mathematics, index of mathematics self-concept, index of mathematics self-efficacy and index of anxiety towards ... mathematics, are based on information gathered from the student questionnaire. In PISA 2012, each index is scaled so that a value of 0 indicates the OECD average and a value of 1 indicates the average standard deviation across OECD countries (see Annex A1 for details on how each index is constructed). Similarly, in PISA 2003, each index was scaled

... READY TO LEARN: STUDENTS’ ENGAGEMENT, DRIVE AND SELF-BELIEFS – VOLUME III © OECD 2013

45

2


so that a value of 0 indicated the OECD average and a value of 1 indicated the average standard deviation across OECD countries. To compare the evolution of these indices over time, the PISA 2012 scale was used and all index values for PISA 2003 were rescaled accordingly. As a result, the values of the indices for 2003 presented in this report differ from those produced in Learning for Tomorrow’s World: First Results from PISA 2003 (OECD, 2004). Also, in PISA 2003, the index of intrinsic motivation to learn mathematics was named index of interest and enjoyment in mathematics. Both the index of intrinsic motivation to learn mathematics of 2012 and the index of interest and enjoyment in mathematics of 2003 are based on the same questionnaire items and can be compared across assessments. * In PISA 2012, several tests were conducted to determine whether the use of country-specific item parameters improved cross-country comparability of indices. For example, simulation studies indicated that using country-specific item parameters in regression models did not lead to improvements in the comparability of indices across countries. During the estimation procedure, an index of differential item functioning (DIF) across countries is produced that can be used to gauge the amount of DIF for each item across countries. If necessary, the impact of DIF on items can then be tackled using country-specific item parameters. However, simulation studies have shown that introducing country-specific item parameters for DIF items has a negligible impact on the regression coefficients in a two-level regression (students within countries) of background variables (with and without country-specific items) on cognitive scores in mathematics, reading and science.

Sources: Bempechat, J., N.V. Jimenez and B.A. Boulay (2002), “Cultural-Cognitive Issues in Academic Achievement: new Directions for Cross-national Research”, in A.C. Porter and A. Gamoran (eds.), Methodological Advances in Cross-National Surveys of Educational Achievement, National Academic Press, Washington, DC. Buckley, J. (2009), “Cross-national response styles in international educational assessments: Evidence from PISA 2006”, Department of Humanities and Social Sciences in the Professions, Steinhardt School of Culture, Education, and Human Development, New York University, New York. King, G. and J. Wand (2007), “Comparing incomparable survey responses: New tools for anching vignettes”, Political Analysis, Vol. 15, pp. 46-66. OECD (2004), Learning for Tomorrow’s World: First Results from PISA 2003, PISA, OECD Publishing. OECD (forthcoming), PISA 2012 Technical Report, PISA, OECD Publishing.

A sense of belonging is not particularly associated with one gender or another: in 20 of the 65 countries and economies that took part in PISA 2012, girls tend to have a stronger sense of belonging than boys while in 13 countries and economies, boys have the stronger sense of belonging (Table III.2.3d). In general socio-economically advantaged students have a stronger sense of belonging than socio-economically disadvantaged students: in 54 countries and economies socio-economically advantaged students reported a stronger sense of belonging, and the difference is particularly large in Liechtenstein, Iceland, France and Lithuania (Table III.2.7b). Socio-economically disadvantaged students are less likely than advantaged students to feel like they belong at school, are more likely to feel like outsiders, and are less likely to feel happy and satisfied with their school. On average, across OECD countries, students in the bottom quarter of the PISA index of economic, social and cultural status have values on the index of sense of belonging that are one quarter of a standard deviation lower than students in the top quarter of the PISA index of economic, social and cultural status (Table III.2.3c). Do 15-year-old students who perform at high levels enjoy greater social acceptance and find it easier to integrate with their peers at school than students who perform less well, or do they suffer a “social penalty” because of their high achievement? The association between academic success and social acceptance may differ for boys and girls, and across socio-economic and ethnic groups (Horner, 1972; Ogbu and Simons, 1998; Spencer and Harpalani, 2008; FullerRowell and Doan, 2010). For example, if doing well at school is not valued among boys, then a boy who is successful academically may be rejected by his male peers and his sense of belonging at school will then weaken (Fuller-Rowell and Doan, 2010; Steele, 1997; Steele, 1998; Davies, Spencer and Steele, 2005). As Figure III.2.14 indicates, students who reported, for example, feeling happy at school, finding it easy to make friends at school and not feeling lonely at school perform better in mathematics than students who reported having less of a sense of belonging (Table III.2.3e). The blue bar in Figure III.2.14 represents the estimated difference in mathematics performance that is associated with a difference of one unit in the index of sense of belonging. This difference corresponds roughly to the variation in the sense of belonging that can be expected between the average student in OECD countries and a student whose sense of belonging places him or her among a group of students with a strong sense of belonging. Only 16.5% of students, on average in OECD countries, reported a stronger sense of belonging than this student (Box III.2.3).

46



2

• Figure III.2.14 • Relationship between sense of belonging and mathematics performance Average student 10th percentile (lowest-achieving students) 90th percentile (highest-achieving students) Score point difference in mathematics, associated with one unit of the index of sense of belonging

25 20 15 10 5 0 -5 -10

Thailand Lithuania Qatar Korea Bulgaria France Liechtenstein Hungary Luxembourg Jordan Slovak Republic Australia Portugal United Arab Emirates Switzerland Czech Republic Indonesia Iceland Romania Peru Belgium Austria Hong Kong-China Estonia Netherlands Argentina Colombia Slovenia OECD average Norway Sweden Mexico Kazakhstan Denmark Tunisia United States United Kingdom Germany Finland Singapore Canada Malaysia Chile Russian Federation Spain Shanghai-China Greece New Zealand Turkey Croatia Japan Israel Costa Rica Brazil Uruguay Chinese Taipei Ireland Viet Nam Serbia Macao-China Italy Poland Latvia Albania Montenegro

-15

Note: Differences that are statistically significant at the 5% level (p < 0.05) are marked in a darker tone. Countries and economies are ranked in descending order of the average score-point difference in mathematics that is associated with one unit of the index of sense of belonging. Source: OECD, PISA 2012 Database, Table III.2.3e. 1 2 http://dx.doi.org/10.1787/888932963806

A sense of belonging is not highly associated with mathematics performance. On average across OECD countries, a difference of one unit in the index of sense of belonging corresponds to a difference of 7 score points in mathematics (Table III.2.3d). In 16 countries and economies, the difference in mathematics performance that is associated with students’ sense of belonging is 10 points or more; in Thailand, Lithuania, Qatar, Korea, Bulgaria and France the gap is somewhat wider, at 15 score points or more. In 18 countries and economies a sense of belonging is not associated with mathematics performance; in all countries and economies the variation in performance associated with a difference of one unit in the index of sense of belonging is smaller than 20 points. Across OECD countries, less than 1% of the variation in students’ mathematics performance can be explained by differences in students’ sense of belonging, and the explained variation in mathematics performance is lower than 5% in all countries and economies. Research examining the association between academic achievement and social acceptance generally confirms a positive circular relationship: social acceptance leads to higher levels of academic achievement, and high levels of academic achievement lead to greater social acceptance (Chen, Rubin and Li, 1997; Wentzel, 1991; Wentzel, 2005; Wentzel, Donlan and Morrison, 2012). However, the link between social acceptance and achievement is likely to differ significantly across countries, depending on whether teenagers value high academic achievement. In some countries, academic achievement is considered socially desirable among teenagers; in others, academic achievement is not a factor in social acceptance among peers, and sometimes it is even sanctioned, particularly in groups of students who do not do well in school or feel marginalised from participation in school (Fordham and Ogbu, 1986; Ogbu, 2003). Results presented in Figure III.2.14 indicate that the relationship between a sense of belonging and mathematics performance is similar regardless of how well students perform. In 22 countries and economies, a sense of belonging is not associated with performance either at the top or at the bottom of the performance distribution. In 15 countries and economies there are differences in performance associated with a sense of belonging among the lowest-achieving students but not among the highest-achieving students; while in Estonia, Sweden, Chile, Hong Kong-China, Montenegro and Colombia, there are READY TO LEARN: STUDENTS’ ENGAGEMENT, DRIVE AND SELF-BELIEFS – VOLUME III © OECD 2013

47

2


differences in performance associated with a sense of belonging among the highest-achieving students but not among the lowest-achieving students. However, in no country is that performance difference large (Table III.2.3e). Overall, the relationship between students’ sense of belonging and their mathematics performance was weak in 2003 in all countries and economies, and remained weak in 2012 (Table III.2.9).

Box III.2.3. The association between students’ dispositions, behaviours and self‑beliefs and mathematics performance Results presented in the chapter on the relationship between students’ engagement with and at school, drive and motivation and self-beliefs and mathematics performance can be used to answer two main policy issues: How strong is the association between mathematics performance and students’ engagement, drive and self-beliefs? Two indicators can be used to answer this question: the slope and the inter-quartile range. The slope represents the score-point difference that is associated with a change of one unit in students’ engagement, drive and self-beliefs. • If this number is low, on average, little or no differences are observed in the mathematics performance of students with different levels of engagement, drive and self-beliefs. Students whose engagement, drive and self-beliefs are similar to those of the average student in an OECD country (index value of 0) show similar performance in mathematics as students who are one standard deviation above the average OECD student in their engagement, drive and self-beliefs (index value of 1). • If this number is high and positive, large differences are observed in the mathematics performance of students with different engagement, drive and self-beliefs. Students whose engagement, drive and self-beliefs are similar to those of the average OECD student (index value of 0) score lower in mathematics score than students who are one standard deviation above the average OECD student in their engagement, drive and self-beliefs (index value of 1). The inter-quartile range represents the difference between the students with the highest and those with the lowest engagement, drive and self-beliefs (i.e. those in the top and bottom quartiles of these indicators within each country). This indicator shows the magnitude of the inequalities in mathematics performance between “enthusiastic” and “unenthusiastic” learners in different countries. Are engagement, drive and self-beliefs good predictors of performance? Identifying the proportion of the variation in student performance that is accounted for by engagement, drive and self-beliefs, the “explained variance”, helps to answer this question. • If this number is low, knowing how engaged students are, their drive and self-beliefs tells very little about their mathematics performance. • If this number is high, by knowing student engagement, drive and self-beliefs one can predict students’ mathematics performance relatively well.

ATTITUDES TOWARDS SCHOOL Students’ attitudes towards school can be influenced by their parents, their teachers, their peers and the atmosphere at school. PISA 2012 sought to find out whether 15-year-olds feel that what they have learned in school is useful for them, both in the immediate and for their future. Students who took part in PISA 2012 were asked to report whether they strongly disagreed, disagreed, agreed or strongly agreed that school has done little to prepare them for adult life when they leave school; that school has been a waste of time; that school has helped to give them confidence to make decisions; and that school has taught them things that could be useful in a job. In addition to these questions, which had also been asked in the PISA 2003 assessment, students participating in the 2012 assessment were also asked to report whether they agree that trying hard at school will help them get a good job, will help them get into a good ,2 whether they enjoy receiving good grades,3 and whether trying hard at school is important. As Figures III.2.15 and III.2.16 and Table III.2.4a illustrate, most students believe that school is useful. Across OECD countries, around nine out of ten students reported that they do not think school has been a waste of time (88%) and

48



2

that they think that school has taught them things that could be useful in a job (87%). Some 71% of students think that school has prepared them for adult life, and 77% believe that school has helped to give them confidence to make decisions. Similarly, Table III.2.5a indicates that 94% believe that trying hard at school will help them get into a good college, and 91% believe that trying hard at school will help them get a good job. However, while students generally reported positive attitudes towards school, their perceptions vary considerably across countries. For example, over 90% of students in Kazakhstan, Albania, Thailand, Peru, Viet Nam, Mexico, Colombia, Indonesia, the Russian Federation and Malaysia think that school has helped to give them confidence to make decisions, while fewer than 70% of students in Japan, Luxembourg, Norway, Korea, Germany, the Netherlands and Israel think so. In 47 of 65 participating countries and economies, girls tended to report more positive attitudes towards school than boys, and in no country or economy did boys report more positive attitudes towards school than girls (Table III.2.4d). Student responses to questions about whether they believe that school has done little to prepare them for adult life, that school has been a waste of time, that it has given them confidence to make decisions, or that it has taught them things that could be useful in a job were used to create the index of attitudes towards school (learning outcomes). The index was standardised to have a mean of 0 and a standard deviation of 1 across OECD countries. Similarly, student responses to questions about whether they believe that trying hard at school will help them get a good job, that trying hard at school will help them get into a good , that they enjoy getting good , and that trying hard at school is important were used to create the index of attitudes towards school (learning activities). The index was standardised to have a mean of 0 and a standard deviation of 1 across OECD countries. The standardisation procedure applied to the indices means that positive values on the index indicate students who have more positive attitudes than the average student in OECD countries, while negative values indicate students who reported less positive attitudes towards school than the average student in OECD countries. Students’ attitudes towards school are not highly associated with mathematics performance. On average across OECD countries, a difference of one unit in the index of attitudes towards school (learning outcomes) as well as a difference of one unit in the index of attitudes towards school (learning activities) correspond to a difference of 9 score points in mathematics (Tables III.2.4d and III.2.5d). In 29 countries and economies, the difference in mathematics performance that is associated with students’ attitudes towards school (learning outcomes) is 10 points or more; in Qatar, Iceland and Australia the gap is 20 score points or more. In 13 countries and economies students’ attitudes towards school (learning outcomes) are not associated with mathematics performance, while in Viet Nam and Turkey the relationship is • Figure III.2.15 • Students’ attitudes towards school: Learning outcomes Percentage of students who reported “agree” or “strongly agree” (a) or who reported “disagree” or “strongly disagree” (b) with the following statements: % 100

80

60

40

20

0 School has done little to prepare me for adult life when I leave school b

School has been a waste of time b

School has helped give me confidence to make decisions a

School has taught me things which could be useful in a job a



49

2


negative. Across OECD countries, around 2% of the variation in students’ mathematics performance can be explained by differences in students’ attitudes towards school (learning outcomes), and the explained variation in mathematics performance is less than 5% in all countries and economies except Qatar, Iceland and Australia, where the explained variation is between 5% and 10%. • Figure III.2.16 • Students’ attitudes towards school: Learning activities Percentage of students across OECD countries who reported that they “agree” or “strongly agree” with the following statements: % 100

80

60

40

20

0 Trying hard at school will help me get a good job

Trying hard at school will help me get into a good

I enjoy receiving good

Trying hard at school is important


Not all changes in students’ attitudes towards school between 2003 and 2012 were positive. While more students reported that school has helped to give them the confidence to make decisions, on average across OECD countries, more students also reported that school is a waste of time. Indeed, in 2012 in the Slovak Republic, Thailand, Tunisia and Poland the share of students who reported that school has been a waste of time increased by more than ten percentage points since 2003, and in Thailand and Tunisia the share of students who reported that school has done little to prepare them for adult life increased by around 20 percentage points during the period. By contrast, attitudes towards school improved in Luxembourg, Austria, Spain, Liechtenstein, Hungary and Japan. In all these countries and economies, the index of attitudes towards school improved by more than 0.1 units between 2003 and 2012. Most notably, students in Japan hold significantly more positive attitudes towards school, across all PISA measures of attitudes towards school, than they did in 2003. Students in Japan in 2012 were over ten percentage points more likely than their counterparts in 2003 to report that school has taught them things that could be useful in a job or that school has given them confidence to make decisions, and they were ten percentage points less likely than their 2003 counterparts to report that school has done little to prepare them for adult life (Figure III.2.17). Students’ attitudes towards school tended to improve the most in countries and economies that also saw improvements in students’ intrinsic and instrumental motivation to learn mathematics (correlations at the country level of 0.4 and 0.5, respectively, Table III.4.10). The relationship between students’ attitudes towards school and their mathematics performance was weak in 2003 and remained weak in 2012 in all countries and economies (Table III.2.9). Do differences in the relationship between dispositions, behaviours and self-beliefs and performance among the highest-achieving and lowest-achieving students reflect the influence of other factors? Previous sections in the chapter describe the relationship between engagement with and at school and mathematics performance. The findings indicate that the relationships between a sense of belonging and mathematics performance estimated for the average student adequately represent what happens for students at all levels of proficiency. However,

50



2

associations estimated for the average student do not fully represent the associations between lack of punctuality and truancy, on the one hand, and mathematics performance, on the other, among highest- and lowest-achieving students. For example, the relationship between arriving late for school and mathematics performance is strongest among the lowest-achieving students. But among these students, there are also differences in performance associated with gender and in socio-economic status. Do these factors affect the relationship between students’ lack of punctuality and mathematics performance? • Figure III.2.17 • Change between 2003 and 2012 in students’ attitudes towards school (learning outcomes) Change in the mean index of attitudes towards school (learning outcomes) between 2003 and 2012

0.4 0.3 0.2 0.1 0 -0.1 -0.2 -0.3 -0.4 -0.5

Tunisia

Slovak Republic

Brazil

Poland

Thailand

Greece

Indonesia

Latvia

Czech Republic

Sweden

Netherlands

Uruguay

Australia

Norway

Macao-China

Denmark

France

Mexico

Turkey

Finland

Portugal

OECD average 2003

Ireland

New Zealand

Canada

Switzerland

Germany

Russian Federation

United States

Italy

Iceland

Korea

Belgium

Luxembourg

Hong Kong-China

Spain

Austria

Liechtenstein

Japan

Hungary

-0.6

Notes: Statistically significant changes between PISA 2003 and 2012 are marked in a darker tone. The figure shows only countries/economies with comparable data in PISA 2003 and PISA 2012. OECD average compares only OECD countries with comparable indices of attitudes towards school since 2003. Countries and economies are ranked in descending order of the change in the mean index of attitudes towards school between PISA 2003 and PISA 2012. Source: OECD, PISA 2012 Database, Table III.2.4e. 1 2 http://dx.doi.org/10.1787/888932963806

In order to examine whether the results presented in previous sections of the chapter reflect the different composition of highest-achieving and lowest-achieving students, Tables III.2.1c, III.2.2c and III.2.3e illustrate two sets of models. The first set, which is used in the previous sections, reports results for arriving late, skipping classes and days of school and for sense of belonging as the only independent variable. The second set reports results from models that further control for students’ socio-economic status and gender. Therefore, results presented in Tables III.2.1c, III.2.2c and III.2.3e represent the performance gap that is associated with students’ engagement with and at school at different points of the performance distribution between students of similar socio-economic status and of the same gender. Table III.2.1c shows results of an analysis of the association between arriving late for school and mathematics performance among the highest-achieving and lowest-achieving students, and how this relationship changes when controlling for students’ socio-economic status and gender. Among the lowest-achieving students, when socio-economic status and gender are taken into account, the strength of the association is reduced; but in most countries this reduction is small: on average across OECD countries, arriving late for school is associated with a 31-point lower score in mathematics. When comparing performance differences among students with similar socio-economic status and of the same gender, the performance difference is 28 points. In Chinese Taipei, France, Singapore and Hungary accounting for socio-economic status and gender reduces by 10 points or more the association between arriving late at school and mathematics performance observed at the bottom of the performance distribution. Table III.2.1c also reveals that, in the majority of countries, controlling for socio-economic differences and for gender has little impact on the association between arriving late for school and mathematics performance among the highest-achieving students.


51

2


On average across OECD countries, the change in mathematics performance associated with arriving late for school among the highest-achieving students is 20 score points, when not controlling for socio-economic status and gender, and 21 points when controlling for these factors. However, in some countries there are notable differences. For example, in Poland, when not controlling for gender and socio-economic status, the performance difference associated with arriving late for school appears to be similar among the lowest- and highest-achieving students (14 points and 16 points, respectively). But when gender and socio-economic status are taken into account, the relationship becomes much stronger among the highest-achieving students: arriving late for school is associated with a drop of 27 score points in mathematics among the highest-achieving students, but a drop of only 16 points among the lowest-achieving students. In Germany, arriving late for school is associated with poorer performance among the lowest-achieving but not among the highest-achieving students before students’ gender and socio-economic status are taken into account. However, when comparing students of the same gender and the same socio-economic status, no performance difference is observed among the lowest-achieving students, but arriving late for school is associated with a 13 score-point drop among the highest-achieving students. Differences in the association between students’ engagement with school across the performance distribution and differences in how estimated associations can vary when controlling for socio-economic status and gender suggest that policy interventions need to be sensitive to the individual student; a one-size-fits-all approach is not appropriate. The relationships that exist among gender, socio-economic status, mathematics performance and students’ engagement with and at school mean that outliers pull average estimates in different directions. Chapter 7 in this volume attempts to disentangle some of the great variety of circumstances in which 15-year-olds are engaged with school, learning and mathematics in order to inform the development of more targeted approaches to education policy. One issue to address is the over-representation of socio-economically disadvantaged students among the lowest-achieving students; a second is the under-performance of girls among the highest-achieving students in mathematics.

Notes 1. Results presented in Chapters 2, 3, 4 and 7 on the association between different indicators of engagement with and at school, drive, motivation and self-beliefs at the top and the bottom of the conditional performance distribution were estimated using quantile regression methods (Koenker and Bassett, 1978; Koenker and Hallock, 2001). 2. The term “college” was adapted in the questionnaires so that it reflected country-specific denominations. 3. The term “grades” was adapted in the questionnaires so that it reflected country-specific denominations.

References Baker, M.L., J.N. Sigmon and M.E. Nugent (2001), “Truancy Reduction: Keeping Students in School”, Juvenile Justice Bulletin, Office of Juvenile Justice and Delinquency Prevention, Washington, D.C. Barber, B.L., M.R. Stone and J.S. Eccles (2010), “Protect, prepare, support, and engage”, in J.L. Meece and J.S. Eccles (eds.), Handbook of Research on Schools, Schooling, and Human Development, pp. 366-378, Routledge, New York. Baumeister, R. and M.R. Leary (1995), “The need to belong: Desire for interpersonal attachments as a fundamental human motivation”, Psychological Bulletin, Vol. 117, pp. 497–529. Baumert, J., J. Nagy and R. Lehmann (2012), “Cumulative Advantages and the Emergence of Social and Ethnic Inequality: Matthew Effects in Reading and Mathematics Development Within Elementary Schools?”, Child Development, 83(4), pp. 1347-1367. Bempechat, J., N.V. Jimenez and B.A. Boulay (2002), “Cultural-Cognitive Issues in Academic Achievement: new Directions for Cross-national Research”, in A.C. Porter and A. Gamoran (eds.), Methodological Advances in Cross-National Surveys of Educational Achievement, National Academic Press, Washington, D.C. Bennett, K.J. and D.R. Offord (2001), “Conduct disorder: Can it be prevented?”, Current Opinion in Psychiatry, 14(4), pp. 333–337. Bonell, C., A. Fletcher and J. McCambridge (2007), “Improving school ethos may reduce substance misuse and teenage pregnancy”, British Medical Journal, Vol. 334, No. 7594, pp. 614–616.

52



2

Buckley, J. (2009), “Cross-national response styles in international educational assessments: Evidence from PISA 2006”, Department of Humanities and Social Sciences in the Professions, Steinhardt School of Culture, Education, and Human Development, New York University, New York. Chen, X., K.H. Rubin and D. Li (1997), “Relation between academic achievement and social adjustment: Evidence from Chinese children”, Developmental Psychology, 33, pp. 518–525. Christenson, S.L., A.L. Reschly and C. Wylie (eds.) (2012), Handbook of Student Engagement, Springer, New York. Davies, P.G., Spencer, S.J. and Steele, C.M. (2005), “Clearing the air: Identity safety moderates the effects of stereotype threat on women’s leadership aspirations”, Journal of Personality and Social Psychology, 88, pp. 276–287. Due, P. et al. (2003), “Socioeconomic health inequalities among nationally representative sample of Danish adolescents: The role of different types of social relations”, Journal of Epidemiology and Community Health, 57, pp. 692-698. Eisenberger, N.I. and M.D. Lieberman (2006), “An experimental study of shared sensitivity to physical pain and social rejection”, Pain, 126, pp. 132-138. Eisenberger, N.I., M.D. Lieberman and K.D. Williams (2003), “Does rejection hurt? An fMRI study of social exclusion”, Science, 302, pp. 290-292. Finn, J. (1989), “Withdrawing from school”, Review of Educational Research, 59(2), pp. 117–142. Fordham, A. and J.U. Ogbu (1986), “Black students’ school success: Coping with the ‘burden of acting white’”, The Urban Review, 18, pp. 176-206. Fredricks, J.A., P.C. Blumenfeld and A.H. Paris (2004), “School engagement: Potential of the Concept, State of the evidence”, Review of Educational Research, Vol. 74, pp. 59-109. Fredricks, J.A. and J.S. Eccles (2006), “Is extracurricular participation associated with beneficial outcomes? Concurrent and longitudinal relations”, Developmental Psychology, Vol. 42, No. 4, pp. 698-713. Fuller-Rowell, T.E. and S.N. Doan (2010), “The social costs of academic success across ethnic groups”, Child Development, Vol. 81, pp. 1696-1713. Griffiths, A-J. et al. (2012), “The relations of adolescent student engagement with troubling and high-risk behaviors”, in S.L. Christenson, A.L. Reschly and C. Wylie (eds.), Handbook of Student Engagement, pp. 563-584, Springer, New York. Harter, S. (1999), The Construction of the Self: A Developmental Perspective, Guildford, New York. Horner, M.S. (1972), “Toward an understanding of achievement-related conflicts in women”, Journal of Social Issues, Vol. 28, No. 2, pp. 157-175. Jenkins, P.H. (1995), “School delinquency and school commitment”, Sociology of Education, Vol. 68, pp. 221-239. Juvonen, J., G. Espinoza and C. Knifsend (2012), “The role of peer relationships in student academic and extracurricular engagement”, in S.L. Christenson, A.L. Reschly and C. Wylie (eds.), Handbook of Student Engagement, Springer, New York, pp. 387-402. King, G. and J. Wand (2007), “Comparing incomparable survey responses: New tools for anching vignettes”, Political Analysis, 15, pp. 46-66. Koenker, R. and G. Bassett (1978), “Regression quantiles”, Econometrica, 46(1), pp. 33-50. Koenker, R. and K.F. Hallock (2001), “Quantile regression”, Journal of Economic Perspectives, 15(4), pp. 143-156. Kross E. et al. (2011), “Social rejection shares somatosensory representations with physical pain”, Proceedings of the National Academy of Sciences of the United States of America, 108, pp. 6270-6275. Ladd, G.W. et al. (2012), “Classroom peer relations and children’s social and scholastic development: Risk factors and resources”, in A.M. Ryan and G.W. Ladd (eds.), Peer Relationships and Adjustment at School, Information Age Press, Charlotte, North Carolina, pp. 11-49. Lee, V.E. and D.T. Burkam (2003), “Dropping out of High-School: The Role of School-Organization and Structure”, American Educational Research Journal, Vol. 40, No. 2, pp. 353-393. McCluskey, C.P., T.S. Bynum and J.W. Patchin (2004), “Reducing chronic absenteeism: An assessment of an early truancy initiative”, Crime and Delinquency, 50(2), pp. 214-234. Nurmi, J.E. et al. (2003), “The Role of Success expectation and task-avoidance in Academic Performance and Satisfaction: three Studies on Antecedents, Consequences and Correlates”, Contemporary Education Psychology, Vol. 28, pp. 59-90. OECD (forthcoming), PISA 2012 Technical Report, PISA, OECD Publishing.


53

2


OECD (2004), Learning for Tomorrow’s World: First Results from PISA 2003, PISA, OECD Publishing. Offord, D.R. and K.J. Bennett (1994), “Conduct disorder: Long-term outcomes and intervention effectiveness”, Journal of the American Academy of Child and Adolescent Psychiatry, 33(8), pp. 1069-1078. Ogbu, J. (2003), “Black American students in an affluent suburb: A study of academic disengagement”, Lawrence Erlbaum, Mahwah, New Jersey. Ogbu, J.U. and H.D. Simons (1998), “Voluntary and involuntary minorities: A cultural-ecological theory of school performance with some implications for education”, Anthropology and Education Quarterly, 29(2), pp. 155-188. Rubin, K.H., W.M. Bukowski and J.G. Parker (2006), “Peer interactions, relationships, and groups”, in W. Damon and N. Eisenberg (eds.), Handbook of Child Psychology, 6th edition, Vol. 3, Wiley, New York, pp. 571-645. Skinner, E.A. and J.R. Pitzer (2012), “Developmental dynamics of student engagement, coping, and everyday resilience”, in S.L. Christenson, A.L. Reschly and C. Wylie (eds.), Handbook of Student Engagement, Springer, New York, pp. 21-44. Spencer, M.B. and V. Harpalani (2008), “What does ‘acting White’ actually mean? Racial identity, adolescent development, and academic achievement among African-American youth”, in J.U. Ogbu (ed.), Minority Status, Collective Identity and Schooling, Lawrence Erlbaum, Mahwah, New Jersey. Steele, C.M. (1997), “A threat in the air: How stereotypes shape intellectual identity and performance”, American Psychologist, 52, pp. 613-629. Steele, C.M. (1998), “Stereotyping and its threats are real”, American Psychologist, 53, pp. 680-681. Valeski, T.N. and D.J. Stipek (2001), “Young children’s feelings about school”, Child Development, 72(4), pp. 1198-1213. Wentzel, K.R. (2005), “Peer relationships, motivation, and academic performance at school”, in A.J. Elliot and C.S. Dweck (eds.), Handbook of Competence and Motivation, Guilford, New York, pp. 279-296. Wentzel, K.R. (1991), “Relations between social competence and academic achievement in early adolescence”, Child Development, 62, pp. 1066-1078. Wentzel, K.R., A. Donlan and D. Morrison (2012), “Peer relationships and social motivational processes”, in A.M. Ryan and G.W. Ladd (eds.), Peer Relationships and Adjustment at School, pp. 79-108, Information Age Press, Charlotte, North Carolina. Wigfield, A. et al. (2006), “Development of achievement motivation”, in W. Damon and N. Eisenberg (eds.), Handbook of Child Psychology, 6th edition, Vol. 3, Wiley, New York, pp. 933-1002.

54
